rare Boss Bf 1, resonance gone mad

[Suburbanite]

MY old Bf-1 (same shape as the roland (pre boss)) has decided to go mad, the resonance is simply gone off the scale and what was once the perfect jet noise has turned in to a whine not dis-similar to a theromin, quite fun, but not what I need,

I've found a schematic, any chance any one has a clue as to why,

any help I would be very gratefull

[52 Bill]

Did you adjust VR4? The resonance control is basically a positive feedback loop. If VR4 should be knocked out of adjustment to a lower resistance it will allow too much feedback and will cause the circuit to oscillate. Read the resonance control adjustment instructions in the service manual and see if that is where the problem lies.

[Mark Hammer]

It's tricky achieving high resonance without introducing oscillation. Consequently, a great many effects that do use a regenerative approach to introducing resonance/emphasis/etc often include a trimpot for fine-tuning how much signal is recirculated. This compensates for the impact of all those component tolerances that may inadvertently end up with the recirculated signal being louder than the input it is mixed with.

If you look at page 4 of the linked-to service manual, you will see that the Resonance pot output/wiper goes to a 100k fixed resistor, then to a 100k trimpot (labelled VR4). You will also see a 150k resistor linking the two halves of IC1 (between pins 7 and 2). In a perfect world, it would really only take an identical 150k fixed resistor to replace the 100k+100k combination from the resonance pot to pin 2, and yield two identical-level signals to mix together when the resonance pot is maxed. But the world is not perfect, so the 100k trimpot is there to create a suitable recirculated-signal level. That may require a combined resistance of more than 150k total...or less.

To adjust it, set the depth for minimum (i.e., no sweep), VR4 to max resistance (100k) and the resonance pot to max. The Manual control can be wherever you want; same with Rate. Now, with a signal source plugged into the BF-1, slowly turn VR4 until you hit the point where oscillation starts. Back away a little, and there you go. Now, even when the resonance control is dimed, there will be no oscillation. If you like living dangerously, leave VR4 right at the edge of oscillation so the potential remains, and use the chassis control to dial back the resonance into tamer territory.

Why did it go wacky? Heaven knows. But warrants considering that trimpots are often not sealed, and oxidation can occur. They're not built terribly well either, and they can get dislodged by bumping. It happens when something is intended for use on the floor where it can be kicked.

Finally, the chassis legending may indicate 9vdc as a power source, but you will note the service manual indicates 12vdc. I have two of these puppies, and I can vouch that they sound better with a 12v supply. The improvement is noted in the width of the sweep. Not A/DA territory, but closer to it than with a 9v supply.

[Suburbanite]

Thanks people, was trying to "tune" the pedal last night but found that when I got rid of the osscilation I all so got rid of most of the effect,

[R.G.]

In 30-year-old pedals, it's good to think about what is happening slowly, in addition to what's happening fast.

If it were my pedal, I would replace every single electrolytic cap in it before doing any more debugging.

[Suburbanite]

In 30-year-old pedals, it's good to think about what is happening slowly, in addition to what's happening fast.

If it were my pedal, I would replace every single electrolytic cap in it before doing any more debugging.

on a scale of 1 to 10, how tricky would it be to (a) find new parts and (b) fit them, I'm happy with an Iron but I do like this pedal a lot, it was made the same month and year as me so I'm quite attached

[Mark Hammer]

The only part that would entail some effort or waiting is the SAD1024 delay chip (which appears to work just fine from what you've said so far). Just about everything else is still widely available, and maybe even cheaper to buy in 2007 dollars than in 1978 dollars.

I'm thinking that perhaps you tweaked the wrong trimpot, or bumped one on the way in. VR4, if misadjusted, will still permit flanging. heck, you can leave it wide open or removed without preventing a flanging effect. But VR2, in particular, needs to be set just right or you DON'T hear any delay signal.

[Suburbanite]

I'll have another go tonight, and let you know

[Suburbanite]

Thank you all very much, pedal is playing nicely again and I'm pretty happy,

[Suburbanite]

volume drop

hey mark, as you have two of these puppies, is there a way around the volume drop when you turn it on, I use a booster pedal and a looper to turn them both on, and that kind of means three pedals rather than one, which might look good but is a pain in the rear to carry around, I was thinking of putting an active volume pot or something.. any clues?

[Mark Hammer]

Well, you first have to consider when it was made, which is essentially well before people started using loop selectors and such. The reference used by the manufacturer was whether "bypass" was the same volume as "effect", not whether a true-bypassed loop-selector would be the same volume as having the effect in circuit.

The other thing to consider is how bypassing is done within the pedal itself. It has been an almost industry-standard practice for damn near 3 decades now that bypassing of effects using a BBD consists largely or almost entirely of simply lifting the "wet" connection to the mixing stage of the effect to negate the effect. The procedure/method is still widely in use by the vast majority of manufacturers.

In the case of the BF-1, they use an interesting amalgam of stompswitch and FETs. Waaaaaaayyyyyy over on the right of the schematic (assuming you rotate it so it reads left to right) you'll see an op-amp stage just above the "Normal/Bypass" switch and below a transistor buffer stage. That op-amp mixes the wet and dry signals together to produce the cancellations. A physical stompswitch is used (instead of what would eventually be a momentary switch and flip-flop circuit) to turn a pair of FETs on and off. the two FETs have different roles for each of the two signal paths. In the one case (wet side) the FET either shunts the signal to ground, cancelling it, or lets it pass undisturbed. In the other case, the FET either presents a lower resistance path in parallel with a 22k resistor, or else (in the off state) allows the effective resistance to be 22k in series with the signal.

You will note that at no time is there ever a change in what happens at the input stage waaaaayyyyyy over at the far left of the schematic. The FET-switching is designed so that effect and bypass are very close in level, but the point of reference is NEVER a straight-wire feedthrough to the output jack.

To adjust the level for balancing bypass and effect with loop selectors, you'll need to do something that affects BOTH wet and dry. That would be either the gain at the input, the gain at the output, or both. Increasing the gain at the input is risky because you don't want to drive the BBD harder than it can take. Increasing the gain at the output is a bit risky in that you can through off the balance of wet and dry if you're not careful. Increasing the gain a wee bit at both ends is probably the wiser route since it does not increase each type/source of risk disproportionately. Given that gain is multiplicative, it also won't take that much adjustment on both ends to achieve the needed increase.

To increase the gain at the front end, increase the value of the 4k7 feedback resistor between pins 6 and 7 of IC1 to 5k1 (the next common value up from 4k7). To increase the gain on the output side, without affecting balance too much, increase the resistor between pins 6 and 7 of IC4. It looks like it's 2k but I can't be sure. Whatever it's current value, don't hike it up by more than 5% or so. That will increase the gain of that stage.

What if the next common value of resistor produces too much gain? If that happens, you can adopt one of two strategies. One would be to replace the 100k resistor to ground on the output with a 100k trimpot and take the output signal from the wiper. Another would be to replace the 470R resistor on the output with another of a slightly higher value (you probably would not want to go higher than 4k7 or 10k).